Programming, Controls & SensorsModule 9 of 13

Computer Vision and Pose Estimation

The most advanced sensing in FRC: cameras that find AprilTags to compute where the robot is, and fusing that with odometry for accurate, drift-free localization.

Computer Vision and Pose Estimation is module 9 of 13 in the Programming, Controls & Sensors guide. It picks up where Closed-Loop Control: PID, Feedforward, and SysId left off and hands over to Worked Examples and Mini-Projects. Its 3 lessons are about 115 minutes of the guide's ~30 hours.

3 lessons~115 min51 lessons in this guide

Start with lesson 1 All of Programming, Controls & Sensors

In this module

3 lessons

01AprilTags and Vision CoprocessorsWhat AprilTags are, why FRC uses them, and the camera systems that detect them.35 min read

02Pose Estimation with Limelight (MegaTag2) and PhotonVisionTurning tag detections into a robust field-relative pose using MegaTag2 and PhotonPoseEstimator.40 min read

03Fusing Vision with OdometryCombining smooth, fast odometry with absolute, drift-free vision using a WPILib pose estimator.40 min read

Previous moduleClosed-Loop Control: PID, Feedforward, and SysId

Next moduleWorked Examples and Mini-Projects

Jargon check

FRC terms used in Computer Vision and Pose Estimation

Longer reads

Articles that go deeper on this

AprilTags in FRC: How Robots See the FieldAprilTags in FRC are the fiducial markers robots use to see the field and know their exact position. How the tags, cameras, and pose estimation actually work.8 min read

FRC Odometry and Pose Estimation: Field-Centric Control with WPILibHow an FRC robot tracks its field position with WPILib: wheel odometry vs pose estimation, gyro heading, fusing AprilTag vision, and field-centric driving.15 min read

All 13 modules in Programming, Controls & Sensors

Department overview